Max, Rolf and others -
I wrote a response to Rolf back when he made his original statement, but
left it in my "outbox" thinking I'd add some things to it, and then never
got around to finishing it. Thanks for spurring me to revisit, Max! And
also thanks for giving the intro to Pascal's wager, which I only allude to
below.
My personal reaction when I first heard about Pascal's wager was twofold:
1. I refuse to believe in a God who would damn me to Hell for not
believing. I'd rather go to Hell. (I decided this when I was eleven. I
thought there was a pretty good likelihood I was damning myself to Hell.
Heavy thing to live with, for an eleven year old.)
2. If I were that kind of God, and the only reason someone believed in me
was as an insurance policy to avoid Hell, that wouldn't be good enough for
me and I'd send them to Hell anyway.
Now here goes with what I wrote a week or so ago...
> (even to speak about "probability that God
>exist" is absurd).
It's not absurd to me.
One of my favorite recreational activities is attempting to formalize
theological speculations in a scientifically meaningful way. I'm in
illustrious company, as Fabio Cozman pointed out in response to this same
thread. (I don't pretend to be the intellectual match of many who have
preceded me in these investigations, but it IS fun to jump into the game.)
Several difficulties arise when trying to do this.
First is the problem of defining God. This is a tough problem, because it
seems to me that by definition, if you can define it, it isn't God. In
other words, to me God is that which lies beyond definition.
On the other hand, how can you talk about the probability that something
you can't define exists?
Well, you can postulate some attributes you think the Universe would have
if there were a God and if there weren't a God, and then you can examine
evidence for the attributes. That is, you can draw the Bayesian network
G ---> A ---> E
where G = "God Exists?" is necessarily a hidden variable, A are attributes
you postulate the world would/wouldn't have if there were/weren't a God,
and E is observable evidence relating to A. You don't need to define God
to judge a probability of an attribute of the world given that God exists
and that God doesn't. You have to bear in mind when you do this, that YOU
HAVE AN INCOMPLETE MODEL!!! But we always have an incomplete model, right?
The question is whether the model is useful. I personally have found
trying to formalize my thinking on these matters to be quite useful. It
has brought me many practical benefits.
So now we come to another difficulty. Any time I've tried to do this and
discuss it with someone, I've discovered enormous disconnects on the
structure of the G ---> A link. People have widely varying opinions,
often very emotionally charged, about what it means to say God exists.
There are usually tacit assumptions that people are unaware they are
making. To some, the existence of God means near-literal truth of
whatever they were taught as children by whatever faith they were raised
in. These I find to be starkly divided between those who have concluded
what they were taught was nonsense if taken literally and therefore this
whole God thing is a bunch of bunk, and those who accept what they were
taught and somehow manage not to be bothered by the apparent contradictions
with everything modern science thinks it knows. Others have managed to
separate out their theology from their childhood instruction /
indoctrination, but there is still wide variation in what they mean by "God
exists." There are also those who believe themselves to be the direct
recipients of "revealed truth" via mystical insight. The problem is, such
people seem to differ wildly on what they think they "know directly" about
the nature of the universe.
Therefore, even to get off the ground, one has to partition both "God
exists" and "God doesn't exist" into different metaphysical positions with
different implications with respect to observable attributes. Note also
that one might also conceive of rival metaphysical stances that have no
differences with respect to potentially observable evidence. That doesn't
mean we can't assign probabilities to them. But of course, if you and I
assign different probabilities to observationally indistinguishable states
of a hidden variable, there is no way to tell who is "right." (Indeed,
neither of us is "right," because all theologies are wrong by definition.)
However (see below for my formulation of theology as a decision problem),
there may still be reasons to prefer one observationally indistinguishable
theological stance over another.
If conducted in a spirit of respectful tolerance, I always learn a lot from
these discussions, and often uncover my own hidden assumptions (this is one
of the most important benefits of serious scientific discussion of
theology).
To me, though, the more meaningful question is not whether God exists, but
the implications of my theology for how I should act in the world. I'm
revealing here my roots in the Western Judeo-Christian tradition, which
brought ethical monotheism to the world. I am also revealing my
philosophical commitment to decision theory, and a discomfort with
attempting to characterize "pure knowledge" in the absence of its
implications for action in the world.
The personal theology I have adopted and continue to evolve as a working
model is the result of a decision analysis. Over the years, I have asked
myself the question of which metaphysical stance has highest utility, where
utility encompasses both my personal utility and benefit to the world
insofar as I am able to affect the world. In other words, I conduct an
ongoing analysis something like Pascal's wager, although a bit more
complex. (Pascal's analysis concerned only whether to live as a good
Christian or as a dissolute hedonist; and his outcomes matrix was pretty
standard Christian.)
Therefore, I don't BELIEVE in God, or in any particular theology. However,
I have adopted the policy of acting as if I believe in a God who satisfies
certain attributes, because I have concluded that belief by myself and a
large proportion of the population in a God approximately like the one I
postulate would lead to more nearly optimal outcomes for myself and for
society as a whole, as compared with rival metaphysical stances I've
considered, under *any* reasonable "ground truth" metaphysic. The analysis
that got me to my theology applies an analysis like Pascal's wager to
derive a "Metaphysical imperative" akin to Kant's Categorical Imperative:
"Adopt, argue for (with respectful tolerance towards dissenters), and
refine with experience, a metaphysical stance such that the world as a
whole would be better off if the metaphysical stance in question were
widely shared." This is a meta-metaphysics that any God I'd care to
believe in would smile upon.
Enough of this. I've done a decision analysis incorporating ethical
considerations, and decided that the time I allocated for a break from the
work I'm doing on a past-deadline paper is up, and I have to get back to
work.
I'd love to continue this discussion with anyone who wants to discuss
particulars of my / your metaphysical stance -- in particular, what are the
properties we think the world would have if this robustly-optimal God
existed? What is this robustly optimal God like and what kind of
relationship does (s)he have with humanity?
Kathy Laskey
